UNI-T UT387B, a tool for locating hidden utilities

For a while, I used a Bosch GBH 4-32 DFR Bosch DMF 10 Zoom. It does a pretty good job of finding metal and wiring, but it belonged to a friend, and it was awkward to ask for it every time. So when I got the chance to review a similar device, I didn't hesitate.


I mentioned the hammer drill model above for a reason: so far, the best hidden-wire detector is a hammer drill :)
We were working at a site once—a large building, about the size of a kindergarten. It was a major renovation, with everything being replaced, and the power had been off for a long time. But one of the installers managed to find a live cable in the wall, in a building with the power off! I don't think any hidden-wire detector could manage that, so now we have a favorite saying: the best hidden-wire detector is a Bosch hammer drill :)
Joking aside, there are hammer drills that combine this feature with the tool, shutting off the motor when they get close to live wiring.

As I mentioned above, I used the Bosch DMF 10 Zoom from time to time, so it'll be easier for me to compare this one with it, especially since they are practically identical in terms of functionality.

But, as always, I'll start with the packaging.
A cardboard box, slightly battered by the post, in UNI-T's signature style, and a black bag in the same style.




The package is very simple:
1. The device itself
2. Bag
3. Manual.

The Bosch also came with two pencils, but those were a bonus related to its design.


The QC stamp is dated November of last year; the manual is in English and Chinese.


Here is the English section of the manual, which also lists the device specifications. I'll list them separately:
Maximum detection depth
Ferrous metals (iron pipes with a minimum diameter of 3 cm) - 80 mm
Non-ferrous metals (aluminum, minimum dimensions 30*2*2 cm) - 80 mm
Live wires - 50 mm
Wood (minimum dimensions 20*16*4 cm) - 20 mm
Operating temperature 0 …40°C
Storage temperature -10 … 50°C
Temperature range for guaranteed measurement accuracy 18 … 28°C
Humidity above 75%
Automatic power-off time 5 min
Battery one 9V alkaline battery (6LR61)
Minimum battery operating voltage about 7.2 V

I got the device specifications from the manufacturer's page. The sensitivity of the two devices is practically the same; the only difference is that Bosch specifies a detection depth of 100 mm for ferrous metal, while UNI-T specifies 80 mm.


The bag is quite handy: the device is held in place by an elastic strap, and there's a pocket at the top. The manual was originally in the pocket, but since it's not particularly useful, you can use it for pencils instead.



The device has the same look as other UNI-T products: the red-and-black color scheme should be familiar to anyone who uses the brand's products.


The design is practically identical to the Bosch Zoom: the sensor is at the top, with the display and a few buttons below.


Actually, there is one significant difference. The Bosch has a hole right through the sensor, making it easier to mark the spots it finds. That's why it came with pencils. In my opinion, it's a small but pretty handy feature.


This one has no hole at the top, so you'll have to mark the spot by eye: find the location, draw a vertical or horizontal line, move the device away, and mark the spot.


Like the Bosch, it has an LED indicator: green means nothing detected, red means something was found. The brightness is about the same as on the Bosch, but the plastic is transparent rather than matte, so the viewing angle matters. In bright sunlight, you won't see the indicator on either the Bosch or the UNI-T.


The display is about the same size as the Bosch's and offers mostly the same functionality. Since both devices have a Zoom function, the display has two scales: the middle one is the main scale, and the one on the right is Zoom.

A little about the Zoom function. The idea is useful, in principle, but I can say I've never needed it, since both devices found everything I needed without it.
It's meant to pinpoint an object's location more accurately. When the device shows a full reading on the middle scale within a 2–3 cm range, press the Zoom button and use the scale on the right to locate it more precisely. Essentially, it's an extension of the main scale. But as I wrote above, the accuracy was good enough for me even without it.
There’s a protective film on the display, but it’s barely noticeable, so I’ve decided to leave it on for now. If I scratch it, I’ll peel it off :)


The Unit has more buttons than the Bosch.
The Bosch had four buttons:
Power
Zoom
Metal
Wood.
The sound was muted by pressing the Metal and Wood buttons simultaneously.
The wiring detection mode could not be turned off.

This one has six buttons:
Power (yellow)
Wood
Metal
Wiring
Zoom (red)
Mute (blue)

Functionally, they do the same thing. The mute function has its own button, and there’s also a button to force wiring detection mode on.


The display doesn’t show much in operating mode. Here’s what it does show:
1. Metal detection
2. Wood detection
3. Hidden wiring detection
4. When the sound is muted, the corresponding icon appears.
5, 6. The device distinguishes ferrous and non-ferrous metals, indicating this with the corresponding icon.


There are four soft, wide “feet” on the bottom. They seem to be designed to help it move more gently over wallpaper, since they’d probably wear out quickly when scanning bare concrete walls :(


The sensor location is marked on top. I’m not quite sure why; it seems more like a statement of fact.
It’s powered by the familiar 9 V battery, just like the Bosch, except that one came with a battery (if I remember correctly). Based on my experience with the Bosch, I’d recommend alkaline batteries. Rechargeable batteries probably won’t work, since the minimum supply voltage is 7.2 V.


And of course, let’s take it apart. There are four self-tapping screws: two in the middle and another two hidden under the bottom “feet.” When I took it apart, I got carried away and pulled off the top “feet” too, but those are just clips.


The two halves are connected by a very short ribbon cable, so I disconnected it right away.


The board with the buttons and display isn’t particularly interesting. Besides, I couldn’t remove it—the clips are too tight.


The main board is more interesting, though. Let’s see if we can figure out what’s going on here.


This is where the coil itself is located. In this case, it acts as the “sensor,” alongside the electronics. The original shape and layout of the traces around the coil are also worth noting.
The build quality is excellent; everything is neat and clean.


There are no components on the back. I think this is to let the board sit more snugly against the case.


At least the metal sensor is inductive. The device presumably works like an oscillator, with the coil in its resonant circuit. When a metal object enters the coil’s field, the oscillation frequency changes, and the device detects that change. When a magnetic material (steel) is brought near, the frequency shifts in one direction; with a non-magnetic material (copper, aluminum), it shifts in the other. The device can therefore indicate what material the detected object is made of.


There’s also a printed coil around the coil wound on the former. By the way, I still haven’t figured out how it detects wooden objects.


On the other side are screens made of traces. They reminded me of an antenna reflector, so I’ll venture a guess that they’re there to narrow the radiation pattern.


There are more subtleties here in how all these screens and coils are positioned relative to one another than in the electronics themselves. It vaguely reminded me of RF engineering and the complex trace patterns on motherboards, although in the latter case they serve a somewhat different purpose.


A little more about the components.
1. CLC1200, an instrumentation amplifier, most likely used as the basis for the oscillator. There’s an LM358 op-amp next to it, apparently acting as a detector. The negative voltage for powering the amplifiers is generated by a CE7660.
2. A couple of small logic ICs were also found on the board; their purpose is unknown.
3. Everything is controlled by a TG842F32 microcontroller, which I’ve never come across before.
4. Voltage regulators for all the electronics. Interestingly, it has a reverse-polarity protection diode (bottom left). Replacing it with a Schottky diode would lower the minimum supply voltage to 6.9–7.0 V from 7.2 V.


Since this isn't a measuring instrument, I ran a few tests and tried to capture them on video.
First, the test setup:
1. Rebar in an interior wall. The rebar is thin, about 4–5 mm, arranged in a mesh and located almost at the wall's surface.
2. Rebar in a load-bearing wall. This rebar is thicker—10–12 mm, if I remember correctly—and sits about 30–40 mm deep. The wall is 160 mm thick.
3. Wiring in a channel about 20 mm deep. This is one of only two places where the old wiring is still in place, so it runs in a channel rather than closer to the surface.
4. An aluminum tube about 5 mm in diameter, a rod of the same diameter made from an unknown alloy, presumably brass, and an 8 mm drill bit.
5. A steel ruler, a piece of copper-clad fiberglass board, and a masonry drill bit with a 12 mm shank.
6, 7. The previous “participants,” tested alongside a 3×0.75 cable, with the ground wire disconnected.


Now for some notes.
A. I couldn't get the claimed 80 mm detection distance. In practice, it was about 50 mm—but I wasn't testing it on a 30 mm tube, just a 12 mm drill-bit shank.
B. It distinguishes aluminum and copper from steel without any trouble. It doesn't respond well to aluminum, though, and the detection distance is significantly reduced.
C. It responds very well to the rod made from the unknown alloy, but identifies it as ferrous metal.
D. It detects wires in the wall and locates them fairly well. It responds far too strongly to a wire lying on the table.
E. I couldn't test for wood in the walls because I didn't have any available. It doesn't respond to a piece of wood on its own, which is understandable.

It may seem like the detector beeps whenever it feels like it. And, basically, that's true: you need to watch the display, where the scale gives you a much clearer picture. The detector starts beeping even when the signal is weak.
To locate wiring more easily, it's best to place your other hand on the same wall. It makes the readings clearer—just a matter of practice.
Yes, the whole process looks like shamanism. You need practice to get a good feel for what's where.
Comparing the Bosch Zoom with the one reviewed here, in my opinion any difference is negligible, since they're in the same class.

If you need to locate absolutely everything with good accuracy, there are other classes of detector—but the price is, to put it mildly, different: roughly an order of magnitude higher.


And finally, here's my red-and-black “family” :)


I won't split this into pros and cons; I'll just give my overall impressions.
In my opinion, this detector isn't much different from the Bosch DMF 10 Zoom. At least, it behaves about the same as the one reviewed here. What's more, I tested it on the same wall where I'd tested the Bosch during my renovation.
With instruments in this class, the results depend more on the operator than on the instrument. But I can say for sure: if it doesn't beep, you can drill. If it does beep (which is almost always), check the display and decide what to do.
Usually, the search isn't about checking one specific spot; it's about scanning the whole area within a radius of 30–40 cm. Rebar is usually closer to the surface and arranged in a mesh, so you just need experience.
Wiring is trickier. If the wall is damp or has lots of small rebar, it'll beep everywhere. The most you can do is put your other hand on the wall, as I mentioned above.

Otherwise, it's a detector like any other: excellent build quality, a sturdy case, reverse-polarity protection, and a dedicated wiring-detection mode (which the Bosch doesn't have). But they skimped on the hole for marking the spot more conveniently.

There are quite a lot of detectors of this type. Some are better, some worse, but at a similar price point they'll all perform about the same.

That's about it. Questions are welcome :)